MCP server for controlling WCP-compatible waveform viewers — Surfer, VaporView, and GTKWave — through the Waveform Control Protocol (WCP).

• Three viewers supported — Surfer, VaporView, GTKWave with automatic protocol handling
• 23 MCP tools covering the full WCP command surface (load, add signals, navigate, markers, etc.)
• Transparent protocol differences — frame format, greeting order, ID types, and JSON-RPC wrapping handled automatically

1. Install and configure:

   pip install .

   Add to your MCP client configuration (e.g., Claude Code settings.json):

   {
     "mcpServers": {
       "wcp": {
         "command": "wcp-mcp",
         "args": []
       }
     }
   }

2. Start a WCP-compatible waveform viewer:

   Surfer — create .surfer/config.toml to autostart the WCP server:

   [wcp]
   autostart = true
   address = "127.0.0.1:54321"

   Then run surfer.

   VaporView — enable WCP in VS Code settings:

   "vaporview.wcp.enabled": true,
   "vaporview.wcp.port": 54322

   Or run VaporView: Start WCP Server from the command palette (Ctrl+Shift+P).

   GTKWave — enable WCP with flags:

   gtkwave --wcp --wcp-port=8765

3. Use the MCP tools:

   connect_viewer(viewer="surfer", port=54321)
   load_waveform(path="/path/to/sim.vcd")
   add_signals(variables=["top.clk", "top.data"])
   zoom_to_fit()
   

MIT